Wear a mask, safe a life.
Last updated Wed, Jul 22 10:13 AM PST.
See “Analyses” tab for past analyses performed.
Decision makers monitor new cases as they’re correlated with deaths 14 days later. Models often predict cumulative deaths, which I feel can hide the recent trends and make it harder to see how much cases are rising or falling. This site is created to show how new cases relates to deaths 14 days later and includes forecasts for deaths. The goal is for this to be a resource. As you see headlines reporting new cases, you can use these plots to put past trends in perspective with current and expected trends.
# Dead/# Confirmed.Predictions were pulled from CDC.gov, which pulls them from CovidHub. The forecast model used is the “Ensemble” model, which averages across various models in a wisdom-of-the-crowd type prediction.
As predictions are reported in cumulative terms, I subtracted cumulative deaths as of the day of prediction and then divided weekly cumulative predictions by 7 to get a daily prediction. E.g., if cumulative predictions are 1000 on a forecast date of 6/1, and 6/8 predictions are 1200, the weekly prediction is 200 and the daily predictions are 200/7=28.5. I plot the daily prediction. Forecast points are on every Saturday. Forecasts happen weekly on Mondays.